Abstract: The electrolyte version of SRK plus association equation of state (eSRK-CPA EoS) was employed to correlate CO₂ solubility in MDEA aqueous solutions. The applied model comprises the classic form of CPA EoS including SRK EoS plus Wertheim association term in addition to MSA theory and Born terms so that the two last terms are responsible for the long-range interactions. A reaction-containing bubble pressure computation technique comprising two nested loops was utilized to model the systems. The internal loop, calculates the liquid phase concentrations via reaction, mass and charge balance equation solving, whereas, the vapor phase concentrations will be obtained in the external one. 470 experimental data were used to correlate binary subsystems and the H₂O + MDEA + CO₂ ternary system. Since, there not exist any binary VLE data for MDEA + CO₂ subsystem, two fitting scenarios were applied. At the first scenario, the binary interaction parameter was assumed equal to zero, while, in second approach the parameter was obtained through ternary system correlation. Both scenarios show very good accuracy in that the Absolute Average Deviation percentages (AAD) obtained were 19.12% and 18.85%, respectively. Also, to show the efficiency of the used model, a comparison between our results and those of the best-known models was made. Finally, having model parameters for H₂S solubility from our previous work[A. Afsharpour, Petroleum Science and Technology 35 (3) (2017) 292-298], simultaneous solubility of CO₂ + H₂S mixtures in MDEA solutions was predicted using the eSRK-CPA EoS with no new optimizable parameters. As the results show, the applied model has a good performance for correlation and prediction of acid gas solubility in a wide range of pressures, temperatures, acid gas loadings, and MDEA concentrations.

Key words: Acid gas solubility, MDEA, Thermodynamic modeling, Phase equilibria, CPA EoS, Prediction